Synchronous Machines 1.0 Introduction One might easily argue that the synchronous generator is the most important component in the power system, since synchronous generators • Are the source of 99% of the MW in most power systems; • Provide frequency regulation and load following; • Are the main source of voltage control; • Are an important source of oscillation damping. For that reason, we will spend the remainder of the course studying this component. EE 303 contains a chapter on synchronous generators (Module G1). Sections 2 and 3 of these notes will be basically a review of this module, except that we will more rigorously develop the smooth rotor model used there. 1 PDF created with pdfFactory trial version www.pdffactory.com Sections 4 and 5 will extend your knowledge of synchronous generators to account for salient pole machines, 2.0 Synchronous Generator Construction The synchronous generator converts mechanical energy from the turbine into electrical energy. The turbine converts some kind of energy (steam, water, wind) into mechanical energy, as illustrated in Fig. 1 [1]. Fig. 1 [1] 2 PDF created with pdfFactory trial version www.pdffactory.com The synchronous generator has two parts: • Stator: carries 3 (3-phase) armature windings, AC, physically displaced from each other by 120 degrees • Rotor: carries field windings, connected to an external DC source via slip rings and brushes or to a revolving DC source via a special brushless configuration. Fig. 2 shows a simplified diagram illustrating the slip-ring connection to the field winding. Rotor winding Brushes +- Stator Stator winding Slip rings Fig. 2 Fig. 3 shows the rotor from a 200 MW steam generator. This is a smooth rotor. 3 PDF created with pdfFactory trial version www.pdffactory.com Fig. 3 Fig. 4 shows the rotor and stator of a hydrogenerator, which uses a salient pole rotor. Fig. 4 4 PDF created with pdfFactory trial version www.pdffactory.com Fig. 5 illustrates the synchronous generator construction for a salient pole machine, with 2 poles. Note that Fig. 5 only represents one “side” of each phase, so as to not crowd the picture too much. In other words, we should also draw the Phase A return conductor 180° away from the Phase A conductor shown in the picture. Likewise for Phases B and C. Phase A ROTOR (field winding) STATOR (armature winding) + N Phase B + DC Voltage The negative terminal for each phase is 180 degrees from the corresponding positive terminal. S + Phase C Fig. 5 5 PDF created with pdfFactory trial version www.pdffactory.com A Two Pole Machine (p=2) Salient Pole Structure Fig. 6 shows just the rotor and stator (but without stator winding) for a salient pole machine with 4 poles. A Four Pole Machine (p=4) (Salient Pole Structure) N S S N Fig. 6 The difference between smooth rotor construction and salient pole rotor construction is illustrated in Fig. 7. Note the air-gap in Fig. 7. Air-gap Fig. 7 6 PDF created with pdfFactory trial version www.pdffactory.com The synchronous generator is so-named because it is only at synchronous speed that it functions properly. We will see why later. For now, we define synchronous speed as the speed for which the induced voltage in the armature (stator) windings is synchronized with (has same frequency as) the network voltage. Denote this as ωe. In North America, ωe=2π(60)= 376.9911≈377rad/sec In Europe, ωe=2π(50)= 314.1593≈314rad/sec On an airplane, ωe=2π(400)= 2513.3≈2513rad/sec The mechanical speed of the rotor is related to the synchronous speed through: ωm = 2 (ωe ) p (1) where both ωm and ωe are given in rad/sec. This may be easier to think of if we write p ωe = (ωm ) 2 7 PDF created with pdfFactory trial version www.pdffactory.com (2) Thus we see that, when p=2, we get one electric cycle for every one mechanical cycle. When p=4, we get two electrical cycles for every one mechanical cycle. If we consider that ωe must be constant from one machine to another, then machines with more poles must rotate more slowly than machines with less. It is common to express ωm in RPM, denoted by N; we may easily derive the conversion from analysis of units: Nm=(ωm rad/sec)*(1 rev/2π rad)*(60sec/min) = (30/π)ωm Substitution of ωm=(2/p) ωe=(2/p)2πf=4πf/p Nm= (30/π)(4πf/p)=120f/p (3) Using (3), we can see variation of Nm with p for f=60 Hz, in Table 1. 8 PDF created with pdfFactory trial version www.pdffactory.com Table 1 No. of Poles (p) ------------------2 4 6 8 10 12 14 16 18 20 24 32 40 Synchronous speed (Ns) ----------------------------3600 1800 1200 900 720 600 514 450 400 360 300 225 180 Because steam-turbines are able to achieve high speeds, and because operation is more efficient at those speeds, most steam turbines are 2 pole, operating at 3600 RPM. At this rotational speed, the surface speed of a 3.5 ft diameter rotor is about 450 mile/hour. Salient poles incur very high mechanical stress and windage losses at this speed and therefore cannot be used. All steam-turbines use smooth rotor construction. 9 PDF created with pdfFactory trial version www.pdffactory.com Because hydro-turbines cannot achieve high speeds, they must use a higher number of poles, e.g., 24 and 32 pole hydro-machines are common. But because salient pole construction is less expensive, all hydromachines use salient pole construction. Fig. 8 illustrates several different constructions for smooth and salient-pole rotors. The red arrows indicate the direction of the flux produced by the field windings. • Synchronous generator Rotor construction Round Rotor Two pole Four Pole Eight Pole Salient Pole ωs = 3600 rpm ωs = 1800 rpm ωs = 900 rpm Fig. 8 10 PDF created with pdfFactory trial version www.pdffactory.com 3.0 Conceptual description The following outlines the conceptual steps associated with production of power in a synchronous generator. 1.DC is supplied to the field winding. 2.If the rotor is stationary, the field winding produces magnetic flux which is strongest radiating outwards from the center of the pole face and diminishes with distance along the air-gap away from the pole face center. Figure 9 illustrates. The left-handfigure plots flux density plotted as a function of angle from the main axis. The right-hand plot shows the main axis and the lines of flux. The angle θ measures the point on the stator from the main axis. In this particular case, we have aligned the main axis with the direct-axis of the rotor. 11 PDF created with pdfFactory trial version www.pdffactory.com Direct rotor axis θ 0 Stator θà Rotor Air B, flux density gap in the air gap Magnetic field lines Fig. 9 3.The turbine rotates the rotor. This produces a rotating magnetic field (text calls it sinusoidal traveling wave) in the air gap, i.e., the plot on the left of Fig. 9 “moves” with time. Figure 10 illustrates, where we see that, for fixed time (just one of the plots), there is sinusoidal variation of flux density with space. Also, if we stand on a single point on the stator (e.g., θ=90°) and measure B as a function of time, we see that for fixed space (the vertical dotted line at 90°, and the red eye on the pictures to the right), there is sinusoidal variation of flux density w/time. 12 PDF created with pdfFactory trial version www.pdffactory.com θ N θà 0 θ θà N θà θ N θ=90° Fig. 10 4.Given that the stator windings, which run down the stator sides parallel to the length of the generator (see Fig. 5) are fixed on the stator (like the eye of Fig. 10), those conductors will see a time varying flux. Thus, by Faraday’s law, a voltage will be induced in those conductors. a. Because the phase windings are spatially displaced by 120°, then we will get voltages that are time-displaced by 120°. 13 PDF created with pdfFactory trial version www.pdffactory.com b.If the generator terminals are opencircuited, then the amplitude of the voltages are proportional to • Speed • Magnetic field strength And our story ends here if generator terminals are open-circuited. 5.If, however, the phase (armature) windings are connected across a load, then current will flow in each one of them. Each one of these currents will in turn produce a magnetic field. So there will be 4 magnetic fields in the air gap. One from the rotating DC field winding, and one each from the three stationary AC phase windings. 6.The three magnetic fields from the armature windings will each produce flux densities, and the composition of these three flux densities result in a single rotating magnetic field in the air gap. 14 PDF created with pdfFactory trial version www.pdffactory.com 7.This rotating magnetic field from the armature will have the same speed as the rotating magnetic field from the rotor, i.e., these two rotating magnetic fields are in synchronism. 8.The two rotating magnetic fields, that from the rotor and the composite field from the armature, are “locked in,” and as long as they rotate in synchronism, a torque (Torque=P/ωm=Force×radius, where Force is tangential to the rotor surface), is developed. This torque is identical to that which would be developed if two magnetic bars were fixed on the same pivot [2, pg. 171] as shown in Fig 11. In the case of synchronous generator operation, we can think of bar A (the rotor field) as pushing bar B (the armature field), as in Fig. 11a. In the case of synchronous motor operation, we can think of bar B (the armature field) as pulling bar A (the rotor field), as in Fig. 11b. 15 PDF created with pdfFactory trial version www.pdffactory.com S S S Bar B Bar A Bar B Bar A N N N N S Fig 11a Fig 11b Fig. 11 4.0 Analytical model: open circuit voltage We will now associate an analytical model with the previous conceptual description. We begin with Fig. 11 (Fig. 6.1 from text). 16 PDF created with pdfFactory trial version www.pdffactory.com Fig. 11 In this figure, note the following definitions: • θ: the absolute angle between a reference axis (i.e., fixed point on stator) and the center line of the rotor north pole (direct rotor axis); it is the same as θ of Fig. 9 if, in Fig. 9, we assume that the direct rotor axis is aligned with the reference axis. • α: the angle made between the reference axis and some point of interest along the air gap circumference. 17 PDF created with pdfFactory trial version www.pdffactory.com Thus we see that, for any pair of angles θ and α, α-θ gives the angular difference between the centerline of the rotor north pole and the point of interest. We are using two angular measurements in this way in order to address • variation with time as the rotor moves; we will do this using θ (which gives the rotational position of the centerline of the rotor north pole) • variation with space for a given θ; we will do this using α (which gives the rotational position of any point on the stator with respect to θ) We want to describe the flux density, B, in the air gap, due to field current iF only. Assume that maximum air gap flux density, which occurs at the pole center line (α=θ), is Bmax. Assume also that flux density B varies sinusoidally around the air gap (as illustrated in Figs. 9 and 10). Then, for a given θ, 18 PDF created with pdfFactory trial version www.pdffactory.com B (α ) = Bmax cos(α − θ ) (4) Keep in mind that the flux density expressed by eq. (4) represents only the magnetic field from the winding on the rotor. But, you might say, this is a fictitious situation because the currents in the armature windings will also produce a magnetic field in the air gap, and so we cannot really talk about the magnetic field from the rotor winding alone. We may deal with this issue in an effective and forceful way: assume, for the moment, that the phase A, B, and C armature windings are open, i.e., not connected to the grid or to anything else. Then, currents through them must be zero, and if currents through them are zero, they cannot produce a magnetic field. So we assume that ia=ib=ic=0. 19 PDF created with pdfFactory trial version www.pdffactory.com So what does this leave us to investigate? Even though currents in the phases are zero, voltages are induced in them. So it is these voltages that we want to describe. These voltages are called the open circuit voltages. Consider obtaining the voltage induced in just one wire-turn of the a-phase armature winding. Such a turn is illustrated in Fig. 12 (Fig. 6.2 of the text). We have also drawn a half-cylinder having radius equal to the distance of the air-gap from the rotor center. Fig. 12 20 PDF created with pdfFactory trial version www.pdffactory.com Note in Fig. 12 that the current direction in the coil is assumed to be from the Xterminal (on the right) to the dot-terminal (on the left). With this current direction, a positive flux direction is established using the right-handrule to be upwards. We denote a-phase flux linkages associated with such a directed flux to be λaa’. Our goal, which is to find the voltage induced in this coil of wire, eaa’, can be achieved using Faraday’s Law, which is: dλ aa ' e aa ' = − (5) dt So our job at this point is to express the flux linking the a-phase λaa, which comes entirely from the magnetic field produced by the rotor, as a function of time. 21 PDF created with pdfFactory trial version www.pdffactory.com An aside: The minus sign of eq. (5) expresses Lenz’s Law [3, pp. 27-28], which states that the direction of the voltage in the coil is such that, assuming the coil is the source (as it is when operating as a generator), and the ends are shorted, it will produce current that will cause a flux opposing the original flux change that produced that voltage. Therefore • if flux linkage λaa’ is increasing (originally positive, meaning upwards through the coil a-a’, and then becoming larger), • then the current produced by the induced voltage needs to be set up to provide flux linkage in the downward direction of the coil, • this means the current needs to flow from the terminal a to the terminal a’ • to make this happen across a shorted terminal, the coil would need to be positive at the a’ terminal and negative at the a terminal, as shown in Fig. 13. 22 PDF created with pdfFactory trial version www.pdffactory.com I I Fig. 13 To compute the flux linking with the coil of wire a-a’, we begin by considering the flux passing through the small slice of the cylinder, dα. The amount of flux through this slice, denoted by dφaa’, will be the flux density at the slice, as given by eq. (4), multiplied by the area of that slice, which is (length) × (width) = (l) × (r dα), that is: 23 PDF created with pdfFactory trial version www.pdffactory.com dφaa' = Bmax cos(α − θ )lrdα (6) = lrBmax cos(α − θ )dα We can now integrate eq. (6) about the halfcylinder to obtain the flux passing through it (integrating about a full cylinder will give 0, since we would then pick up flux entering and exiting the cylinder). φaa ' = π /2 ∫ lrB max cos(α − θ )dα −π / 2 π /2 = lrBmax sin(α − θ ) −π / 2 π π = lrBmax sin − θ − sin − − θ 2 2 14 2 43 14 4244 3 − cosθ cos θ (7) = lrBmax (cos(θ ) − − cos(θ )) = 2lrBmax cosθ Define φmax=2lrBmax, and we get φaa ' = ϕ max cosθ (8) which is the same as eq. (6.2) in the text. 24 PDF created with pdfFactory trial version www.pdffactory.com Equation (8) indicates that the flux passing through the coil of wire a-a’ depends only on θ. That is, • given the coil of wire is fixed on the stator, and • given that we know the flux density occurring in the air gap as a result of the rotor winding, • we can determine how much of the flux is actually linking with the coil of wire by simply knowing the rotational position of the centerline of the rotor north pole (θ). But eq. (8) gives us flux, and we need flux linkage. We can get that by just multiplying flux φaa’ by the number of coils of wire N. In the particular case at hand, N=1, but in general, N will be something much higher. Then we obtain: λaa ' = Nφaa ' = Nϕ max cosθ (9) 25 PDF created with pdfFactory trial version www.pdffactory.com Now we need to understand clearly what θ is. It is the centerline of the rotor north pole, BUT, the rotor north pole is rotating! Let’s assume that when the rotor started rotating, it was at θ=θ0, and it is moving at a rotational speed of ω0, then θ = ω0t + θ 0 (10) Substitution of eq. (10) into eq. (9) yields: λaa ' = Nϕ max cos(ω 0 t + θ 0 ) (11) Now, from eq. (5), we have eaa ' = − dλaa' − d ( Nϕ max cos(ω 0t + θ 0 )) (12) = dt dt We get a –sin from differentiating the cos, and thus we get two negatives, resulting in: e aa ' = N ϕ max ω 0 sin (ω 0 t + θ 0 ) (13) Define E max = Nϕ maxω 0 (14) Then eaa ' = Emax sin(ω 0 t + θ 0 ) (15) We can also define the RMS value of eaa’ as 26 PDF created with pdfFactory trial version www.pdffactory.com Eaa ' = Emax 2 (16) which is the magnitude of the generator internal voltage. We have seen internal voltage before, in EE 303, where we denoted it as |Ef|. In EE 303, we found it in the circuit model we used to analyze synchronous machines, which appeared as in Fig. 14. Note that internal voltage is the same as terminal voltage on the condition that Ia=0, i.e., when the terminals are open-circuited. This is the reason why internal voltage is also referred to as open-circuit voltage. Ia jXs Ef Vt Fig. 14 27 PDF created with pdfFactory trial version www.pdffactory.com Zload We learned in EE 303 that internal voltage magnitude is proportional to the field current if. This makes sense here, since by eqs. (14) and (16), we see that Eaa ' Emax Nϕ maxω0 = = 2 2 (17) and with N and ω0 being machine design parameters (and not parameters that can be adjusted once the machine is built), the only parameter affecting internal voltage is φmax, which is entirely controlled by the current in the field winding, if. One last point here: it is useful at times to have an understanding of the phase relationship between the internal voltage and the flux linkages that produced it. Recall eqs. (11) and (13): λaa ' = Nϕ max cos(ω 0 t + θ 0 ) (11) e aa ' = N ϕ max ω 0 sin (ω 0 t + θ 0 ) (13) Using sin(x)=cos(x-π/2), we write (13) as: eaa ' = Nϕ maxω 0 cos(ω 0 t + θ 0 − π / 2) (18) 28 PDF created with pdfFactory trial version www.pdffactory.com Comparing eqs. (11) and (18), we see that the internal voltage lags the flux linkages that produced it by π/2=90° (1/4 turn). This is illustrated by Fig. 15 (same as Fig. E6.1 of Example 6.1). In Fig. 15, the flux linkage phasor is in phase with the direct axis of the rotor. Reference Axis θ0=π/4 Flux Linkage Phasor Λaa’ Reference Axis Internal Voltage Phasor Eaa’ a-phase armature winding Fig. 15 Therefore, the flux linkages phasor is represented by Nφmax jθ 0 Λ aa ' = e = Λ aa ' e jθ 0 (19) 2 and then the internal voltage phasor will be 29 PDF created with pdfFactory trial version www.pdffactory.com E aa ' = Nφmaxω 0 e j (θ 0 −π / 2 ) = E aa ' e j (θ 0 −π / 2 ) (20) 2 Let’s drop the a’ subscript notation from Eaa’, just leaving Ea, so that: Nφmaxω 0 j (θ 0 −π / 2 ) Ea = e = E a e j (θ 0 −π / 2 ) (21) 2 Likewise, we will get similar expressions for the b- and c-phase internal voltages, according to: E b = E a e j (θ 0 −π / 2 − 2π / 3 ) (22) E c = E a e j (θ 0 −π / 2 + 2π / 3 ) (23) 5.0 Armature reaction: one phase winding Armature reaction refers to the influence on the magnetic field in the air gap when the phase windings a, b, and c on the stator are connected across a load. 30 PDF created with pdfFactory trial version www.pdffactory.com Let’s consider a smooth rotor, as given in Fig. 16 (Fig. 6.3 in text). For a smooth rotor, the flux sees constant (& high) permeability path throughout the rotor and stator, with the only exception being the air gap. Reference Axis ● × a-phase armature winding Fig. 16 Two important observations from Fig. 16: 1.Because the armature windings are stationary, and since the rotor is round (and therefore flux path is constant), the way the picture characterizes lines of flux is representative for all time. The only temporal variation will be the magnetic field strength as it pulsates with current. 31 PDF created with pdfFactory trial version www.pdffactory.com 2.The lines of flux made by each side of the a-phase winding will combine in the rotor in such a way so that its maximum strength is along the reference axis, and it varies sinusoidally along the air-gap. Let’s assume that the current in the a-phase winding is given by: ia ( t ) = 2 I a cos(ω 0 t + ∠I a ) (24) where |Ia| is the RMS current magnitude and ∠I a is the angle made by the current phasor enabling proper phase relation with its corresponding voltage phasor, eq. (18), ea ( t ) = Nϕ maxω 0 cos(ω 0 t + θ 0 − π / 2) (18) where ∠Ea = θ 0 − π / 2 . Then the air-gap flux density at the reference axis, where the flux density is maximum, is given by B a ,max ( t ) = Ki a ( t ) (25) 32 PDF created with pdfFactory trial version www.pdffactory.com We emphasize eq. (22) gives flux density at the reference axis only, a fixed point in the air gap corresponding to α=0°. Fig. 17 uses the orange to illustrate what eq. (25) is capturing. The sequence of the 4 pictures, numbered 1,2,3,4, correspond to the sequence-in-time that characterizes flux density seen at the α=0° fixed stator point. 1 ● 2 α=0° ● × 3 × × 4 × ● ● Fig. 17 Because flux minimizes travel in highreluctance paths, all flux lines are directed radially outwards in the airgap. 33 PDF created with pdfFactory trial version www.pdffactory.com Substitution of eq. (24) in eq. (25) results in: Ba ,max ( t ) = K 2 I a cos(ω 0 t + ∠I a ) (26) Studying the pictures of Fig. 17, we observe that if the flux density is maximum at α=0°, then it will be minimum (i.e., a maximum but negative, or radially directed inwards) at α=180°, and it therefore must be zero halfway between these two points, at α=90° and α=270°. Moving from the maximum to the zero-point in either direction, the flux density will decrease, and a similar thing will occur in moving from the minimum to the zero-point in either direction. Figure 18 illustrates. α=0° Radially outward is positive; radially inward is negative. α=90° ● × α=270° α=180 ° Fig. 18 34 PDF created with pdfFactory trial version www.pdffactory.com And so, for a particular time t, we can express the spatial variation of the flux density from the a-phase winding as a sinusoidal function of α, having a value of Ba,max(t) at α=0°. Therefore: Ba (α , t ) = Ba ,max ( t ) cos α (27) Substituting eq. (26) into (27), we obtain: Ba (α , t ) = K 2 I a cos (ω 0 t + ∠I a ) cos α (28) Now define ′ = K 2 Ia Bmax (29) (This B’max is a little different from that implied in the text, page 195, because we have assumed a perfect sinusoidal distribution of flux density in the air gap, whereas the book assumed a square wave distribution and then took the fundamental of that square wave). Substituting (29) into (28) results in: ′ cos(ω 0 t + ∠I a ) cos α Ba (α , t ) = Bmax 35 PDF created with pdfFactory trial version www.pdffactory.com (30) Question: Does eq. (30) characterize a rotating magnetic field? To answer this question, we need to recall precisely what we mean by a rotating magnetic field. Let’s re-examine the rotating magnetic field developed by the rotating rotor winding. We expressed this in eq. (4): B (α ) = Bmax cos(α − θ ) (4) Inserting explicitly the dependence of θ on t: B (α , t ) = Bmax cos(α − θ ( t )) (31) where θ(t)=ω0t+θ0. So there are three attributes: 1.Constant amplitude 2.One sinusoid 3.Argument of sinusoid a function of space and time Equation (30) has two sinusoids with one being a function of time (first one) and the other being a function of space (second one). And employing trig identities can not result in satisfying the above criteria. 36 PDF created with pdfFactory trial version www.pdffactory.com Physically, a rotating magnetic field • maintains a constant amplitude waveform that • continuously moves around the air gap. Equation (30), on the other characterizes a field that • has a time-varying amplitude • which is stationary in the air gap. hand, If you stand at a particular point on the air gap and observe only the field at that point, you will not be able to tell the difference between the two. However, if you cut the stator and spread out the air gap linearly, and observe the entire waveform of both fields, you will find: • the rotating magnetic field moves. • the other one pulsates. 37 PDF created with pdfFactory trial version www.pdffactory.com A good way to think about this is to consider that in the case of the rotating magnetic field, the flux density is never 0 everywhere in the air gap.WS In contrast, every time the current in the aphase winding goes to zero, the flux density of the eq. (30) pulsating field goes to zero everywhere in the air gap. Cool. 6.0 Armature reaction: all phase windings We can go through a similar thought process to obtain the flux density produced by the band c-phase armature windings. The result will be identical to eq. (30), except that • the time-dependent term will be phaseshifted consistent with the phase-shifts of the current, and • the space-dependent term will be phaseshifted consistent with the phase-shifts of the physical windings. 38 PDF created with pdfFactory trial version www.pdffactory.com Therefore, the flux densities from all phases will be: ′ cos (ω 0 t + ∠I a ) cos α Ba (α , t ) = Bmax ′ cos ω 0 t + ∠I a − Bb (α , t ) = Bmax ′ cos ω 0 t + ∠ I a + Bc (α , t ) = Bmax 2π 3 2π 3 cos α − cos α + 2π 3 2π 3 What will the composition of these three flux densities look like? Let’s add them up! ′ cos(ω 0 t + ∠ I a )cos α Babc (α , t ) = Bmax 2π ′ cos ω 0 t + ∠ I a − + Bmax 3 2π ′ + Bmax cos ω 0 t + ∠ I a + 3 2π cos α − 3 2π cos α + 3 Factor out the B’max: ′ cos(ω 0 t + ∠I a ) cos α Babc (α , t ) = Bmax 2π 2π + cos ω 0 t + ∠ I a − cos α − 3 3 2π 2π + cos ω 0 t + ∠ I a + cos α + 3 3 39 PDF created with pdfFactory trial version www.pdffactory.com Now deploy the trig identity cos a cos b = 1 [cos(a − b) + cos(a + b)] 2 to all three terms. ′ Bmax cos (ω 0 t + ∠ I a − α ) + cos (ω 0 t + ∠ I a + α ) 2 2π 2π 2π 2π + cos ω 0 t + ∠ I a − −α + +α − + cos ω 0 t + ∠I a − 3 3 3 3 Babc (α , t ) = 2π 2π 2π 2π +α + + cos ω 0 t + ∠ I a + −α − + cos ω 0 t + ∠I a + 3 3 3 3 Combine all of the terms with π. ′ Bmax 2 cos (ω 0 t + ∠ I a − α ) + cos (ω 0 t + ∠ I a + α ) 4π + cos (ω 0 t + ∠I a − α ) + cos ω 0 t + ∠ I a + α − 3 4π + cos (ω 0 t + ∠I a − α ) + cos ω 0 t + ∠ I a + α + 3 Babc (α , t ) = Now we notice a very interesting thing. The three cos terms on the right side of each row constitute a balanced set (equal-magnitude terms 120° out of phase). They therefore add to zero! So we have ′ Bmax {cos(ω 0 t + ∠I a − α ) 2 + cos (ω 0 t + ∠ I a − α ) Babc (α , t ) = + cos (ω 0 t + ∠ I a − α )} 40 PDF created with pdfFactory trial version www.pdffactory.com But now we can see that all three of the remaining cos terms are identical! So we have: Babc (α , t ) = ′ 3 Bmax cos (ω 0 t + ∠I a − α ) 2 (32) And this is an expression for a rotating magnetic field because it has 1.Constant amplitude 2.One sinusoid 3.Argument of sinusoid a function of space and time Note that the text, in eq. (6.15) gives: Babc (α , t ) = ′ 3 Bmax cos (α − ω 0 t − ∠ I a ) 2 (33) But eq. (32) and (33) are the same, since cos(x)=cos(-x). Comment on angular relations: Consider eq. (33) when t=0: Babc (α , t = 0) = ′ 3 Bmax cos (α − ∠ I a ) 2 41 PDF created with pdfFactory trial version www.pdffactory.com (34) The maximum of this rotating magnetic field occurs when the argument of the cosine function is 0, and this occurs when α = ∠I a . Another way of saying this is, at t=0, the centerline (or north pole) of this rotating magnetic field (from armature reaction) is at an angle ∠I a with respect to the reference axis. Now consider the following equations that we have previously used: θ = ω0t + θ 0 (10) λaa ' = Nϕ max cos(ω 0 t + θ 0 ) (11) eaa ' = Nϕ maxω 0 cos(ω 0 t + θ 0 − π / 2) (18) These equations say the following for angular relations at t=0. • From eq. (10), θ = θ 0 implies that the north pole of the rotor is at θ0. • From eq. (11), the a-phase flux linkage phasor is also at θ0, aligned with the north pole of the rotor. 42 PDF created with pdfFactory trial version www.pdffactory.com From eq. (18), the a-phase induced voltage is at θ0-π/2, 90° behind the rotor and the flux linkage phasor. • And from eq. (34), as mentioned, the north pole of the armature reaction rotating magnetic field is at and angle ∠I a . Figure 19 illustrates. • θ0 ∠I Λa Ea Fig. 19 43 PDF created with pdfFactory trial version www.pdffactory.com 7.0 Reactances and terminal voltage You will recall that when we were developing the induced voltage in the armature windings (the open circuit voltage) from the rotating magnetic field produced by the rotor, we expressed the flux density from the rotor as B (α ) = Bmax cos(α − θ ) (4) and then we integrated to get total flux about the half-cylinder surrounding the rotor via: λ aa ' = N π /2 ∫ lrB max cos(α − θ )dα −π / 2 π /2 = NlrBmax sin(α − θ ) −π / 2 π π = NlrBmax sin − θ − sin − − θ 2 14 42244 3 243 14 − cos θ cos θ = NlrBmax (cos(θ ) − − cos(θ )) = 2 NlrBmax cosθ = φmax cosθ = Nφmax cos(ω 0 t + θ 0 ) (7) 44 PDF created with pdfFactory trial version www.pdffactory.com Finally, we used Faraday’s law to get dλaa ' − d ( Nϕ max cos(ω0 t + θ 0 )) (12) eaa ' = − = dt dt eaa ' = Emax sin(ω 0 t + θ 0 ) (15) The whole point of the above exercise was that the a-phase windings were experiencing flux linkages that varied with time, and this produced a voltage in which we were interested. Now the a-phase windings are seeing additional fields from armature reaction, and as we have shown, the composite of the fields from all three phases is a rotating magnetic field. Therefore the a-phase is experiencing flux linkages that vary with time (in addition to those from the rotor field) that are caused by the rotating magnetic field of armature reaction. 45 PDF created with pdfFactory trial version www.pdffactory.com So let’s define λag as the total air-gap flux linkages seen by coil aa’. Then, as eq. (6.18) states in the text: λ ag = λ aa ' + λ ar (35) where • λaa’ is the flux linkages from the rotating magnetic field of the rotor • λar is the flux linkages from the rotating magnetic field of armature reaction We want to obtain the voltage induced by the time variation in λag. This will be: − dλ ag − dλ aa ' − dλ ar v ag = = + (36) dt dt dt But we have already obtained the first term in eq. (36); it was eaa’ as given by eq. (15). You will recall that we renamed this ea, and that we showed it had a phasor representation, from eq. (21), of E a = E a e j (θ −π / 2 ) (21) So our job is only to obtain the second term in eq. (36). 0 46 PDF created with pdfFactory trial version www.pdffactory.com I will spare you the details of this effort. Suffice it to say that the steps are virtually identical to the steps we took to get eqs. (7), (12), (15), and (21), and are provided in the text on pp. 196-197. But I do want to impress upon you four important ideas that are at the heart of this development. The first two and the last one are articulated by the text, the 3rd one is not. • Although the flux linkages λar are due to all three phase currents, the voltages induced by dλar/dt are given by − dλ ar di a v ar = = − Ls1 (37) dt dt where the influence of the other two phases is absorbed into the inductance Ls1. • The expression (37), in the time-domain, can also be expressed in the phasor domain as: Var = − jω 0 Ls1 I a (38) 47 PDF created with pdfFactory trial version www.pdffactory.com • • Because of Lenz’s law, the voltage induced by the armature reaction flux linkages will be in opposition to the voltage which produced the time varying currents responsible for these time varying flux linkages. This means that this induced voltage will subtract from Ea: Vag = E a − Var = E a − jω 0 Ls 1 I a (39) Eq. (39) is the same as (6.23) in the text: In the ideal world, Vag would be the terminal voltage, but in reality, the machine will also experience two other effects that need to be included: o Flux leakage o Conductor resistance To account for these two effects, we subtract from eq. (39) a drop across an impedance r+jXl, where r is the resistance of the a-phase armature winding, and Xl is a small reactance that accounts for the drop in induced voltage caused by some of the flux leaking. Therefore: 48 PDF created with pdfFactory trial version www.pdffactory.com Va = Vag − ( r + jX l ) I a (40) Substitution of eq. (39) into eq. (40) results in Va = E a − jω 0 Ls 1 I a − ( r + jX l ) I a (41) Define Xs1=jω0Ls1, so that eq. (41) becomes: Va = E a − jX s 1 I a − ( r + jX l ) I a (42) Now rearrange eq. (42) to obtain: Va = E a − j ( X s1 + X l ) I a − rI a (43) Now define the synchronous reactance Xs=Xs1+Xl, so we obtain: Va = E a − jX s I a − rI a (44) Eq. (44) is the same as eq. (6.24) in the text. The text mentions (page 197) that the armature reaction term Xs1 is the dominant term; the leakage reactance Xl is typically only about 10% of Xs1, and r is only 1%. Figure 20 illustrates per-phase circuit of a synchronous generator, and shows the relation between eqs. (43) and (44): 49 PDF created with pdfFactory trial version www.pdffactory.com jXs Ia + Ea - jXsl jXl r + Va - + Vag - Fig. 20 One last thing: the angle by which the phasor Ea leads the phasor Va is typically referred to as the power angle and represented by δm, i.e., δ m = ∠E a − ∠V a (45) If Va is the reference (has 0 degree angle), then δm is the angle of the internal voltage. Eq. (44) provides the instruction manual for drawing a phasor diagram of a synchronous machine. Usually, however, we begin assuming that we know Va and Ia. We assume that Va is the reference (i.e., has an angle of 0 degrees). 50 PDF created with pdfFactory trial version www.pdffactory.com To obtain Ea from Va and Ia, we re-write eq. (44) like this: E a = Va + jX s I a + rI a (46) Now draw the Ea vector for the lagging condition (this means that Ia is lagging Va): Now draw the Ea vector for the leading condition (this means that Ia is leading Va): 51 PDF created with pdfFactory trial version www.pdffactory.com What can you say about the relative magnitude of the field current in the two cases above? Lagging: Leading: What can you say about the strength of the magnetic field produced by the rotor winding, and the var supply, in the two cases above? Lagging: Leading: 8.0 Power for the smooth rotor case The per-phase equivalent circuit of Fig. 20 is simplified to that of Fig. 21: 52 PDF created with pdfFactory trial version www.pdffactory.com Ia jXs + Ea - + Va - Fig. 21 The complex power will be E a − Va S = V I = Va jX s * a a * (47) Assuming that Va is the reference, and with Ea=|Ea|(cosδm+jsinδm), we obtain: E a cos δ m + j E a sin δ m − Va S = Va jX s * (48) Multiplying top and bottom by j we get: j E a cos δ m − E a sin δ m − jVa S = Va − Xs * − j E a cos δ m + E a sin δ m + jVa = Va + Xs 53 PDF created with pdfFactory trial version www.pdffactory.com * Rearranging, we obtain: E a sin δ m − j E a cos δ m + jVa S = Va X s * Taking the conjugate of what is inside the parenthesis: E a sin δ m + j E a cos δ m − jVa S = Va X s Recalling Va has 0 degree angle, we distribute it through: S= Va E a sin δ m + j Va E a cos δ m − j Va 2 Xs Recalling S=Pg+jQg, we can separate the above expression into its real and imaginary parts to obtain: Pg = Qg = Va E a sin δ m (49) Xs Va E a cos δ m − Va Xs 54 PDF created with pdfFactory trial version www.pdffactory.com 2 (50) Equation (50) provides us with an important way to understand the relation between leading and lagging generator operation. The difference between leading and lagging will occur when Qg=0. Qg = 2 Va E a cos δ m − Va =0 Xs 2 Va E a cos δ m − Va = 0 Va E a cos δ m = Va 2 E a cos δ m = Va (51) So, eq. (51) provides the condition corresponding to the boundary between leading and lagging. Draw the corresponding phasor diagram for this boundary, and from that, identify the condition for lagging generator operation and the condition for leading generator operation. 55 PDF created with pdfFactory trial version www.pdffactory.com 56 PDF created with pdfFactory trial version www.pdffactory.com 9.0 Excitation control Moving between lagging and leading condition is performed via control of the generator field current, which produces the field flux φ . Field current control can be done manually, but it is also done automatically via the excitation control system. f The excitation control system is an automatic feedback control having the primary function of maintaining a predetermined terminal voltage by modifying the field current of the synchronous generator based on changes in the terminal voltage. Without excitation control, terminal voltage would fluctuate as a result of changes in Pg or external network conditions. The control is referred to as “negative feedback” because when terminal voltage increases, field current is decreased, and when terminal 57 PDF created with pdfFactory trial version www.pdffactory.com voltage decreases, field current is increased. A simplified block diagram of an excitation control system is shown in Figure G1.9. Fig. 22 Block Diagram of Excitation Control System There are three fundamental components to any excitation system. The main exciter, or more simply, the exciter, is the device that provides the field current for the synchronous generator. The automatic 58 PDF created with pdfFactory trial version www.pdffactory.com voltage regulator (AVR) couples the terminal voltage to the input of the main exciter. The amplifier increases the power of the regulating signal to that required by the exciter. If the amplifier is electromechanical, it is called the pilot exciter or the rotating amplifier. If the amplifier is solid state, it is usually considered as part of the AVR. There are three basic types of excitation systems. These are: • rotating DC commutator • rotating AC alternator • static These are illustrated in Fig. 23, 24, and 25. 59 PDF created with pdfFactory trial version www.pdffactory.com Fig. 23 Rotating DC Commutator Type Excitation System The DC commutator excitation system utilizes a DC generator mounted on the shaft of the synchronous generator to supply the field current. This type of system is no longer used in new facilities because it is slow in response, and because it requires high maintenance slip rings and brushes to couple the exciter output to the field windings. 60 PDF created with pdfFactory trial version www.pdffactory.com Fig. 24 Rotating AC Alternator Type Excitation System The AC alternator excitation system uses an AC alternator with AC to DC rectification to supply the field winding of the synchronous generator. An important advantage over DC commutator systems is that AC alternator systems may be brushless, i.e., they do not use slip rings to couple the exciter to the rotor-mounted field winding. For example, the General Electric Althyrex uses an “inverted” alternator to supply the field voltage through a rectifier. The alternator is inverted in that, unlike the power generator, the field winding is on the stator and the 61 PDF created with pdfFactory trial version www.pdffactory.com armature windings are on the rotor. Therefore the alternator field can be fed directly without the need for slip rings and brushes. Rectification to DC, required by the synchronous generator field, takes place by feeding the alternator three-phase output to a thyristor controlled bridge. The thyristor or silicon controlled rectifier (SCR) is similar to a diode, except that it remains “off” until a control signal is applied to the gate. The device will then conduct until current drops below a certain value or until the voltage across it reverses. Figure 25: Static Type Excitation System 62 PDF created with pdfFactory trial version www.pdffactory.com The third type of excitation system is called a static system because it is composed entirely of solid state circuitry, i.e., it contains no rotating device. The power source for this type of system is a potential and/or a current transformer supplied by the synchronous generator terminals. Threephase power is fed to a rectifier, and the rectified DC output is applied to the synchronous generator field via slip rings and brushes. Static excitation systems are usually less expensive than AC alternator types, and the additional maintenance required by the slip rings and brushes is outweighed by the fact that static excitation systems have no rotating device. [1] http://geothermal.marin.org/GEOpresentation/ [2] A. Fitzgerald, C. Kingsley, and A. Kusko, “Electric Machinery, Processes, Devices, and Systems of Electromechanical Energy Conversion,” 3rd edition, 1971, McGraw Hill. [3] S. Chapman, “Electric Machinery Fundamentals,” 1985, McGraw-Hill. 63 PDF created with pdfFactory trial version www.pdffactory.com